Comparing Different Multilevel Bootstrapping Methods

Data

Here I use a synthetic dataset (pop_syn) bundled with the package, which mimics the structure of the example data in Hox (2010, p. 17). The synthetic data was generated from parameters estimated from the original popular dataset.

library(dplyr)
library(lme4)
library(boot)
library(bootmlm)

For the interest of time, I will select only 30 schools and a few students in each school:

set.seed(85957)
pop_sub <- pop_syn |> filter(school %in% sample(unique(school), 30)) |>
  group_by(school) |> sample_frac(size = .25) |> ungroup()

To illustrate doing bootstrapping to obtain standard error (SE) and confidence interval (CI) for the intraclass correlation (ICC) of the variable popular, we first fit an intercept only model:

m0 <- lmer(popular ~ (1 | school), data = pop_sub)
summary(m0)

## Linear mixed model fit by REML ['lmerMod']
## Formula: popular ~ (1 | school)
##    Data: pop_sub
## 
## REML criterion at convergence: 441.8
## 
## Scaled residuals: 
##      Min       1Q   Median       3Q      Max 
## -1.84638 -0.60538 -0.03404  0.67043  2.42008 
## 
## Random effects:
##  Groups   Name        Variance Std.Dev.
##  school   (Intercept) 0.9692   0.9845  
##  Residual             0.7110   0.8432  
## Number of obs: 152, groups:  school, 30
## 
## Fixed effects:
##             Estimate Std. Error t value
## (Intercept)   5.2634     0.1925   27.34

The intraclass correlation is defined as $\rho = \frac{\tau}{\tau + \sigma^2} = \frac{1}{1 + \sigma^2 / \tau} = \frac{1}{1 + \theta^{-2}},$ where $\theta = \sqrt{\tau} / \sigma$ is the relative cholesky factor for the random intercept term used in lme4. Therefore, we can estimate the ICC as:

(icc0 <- 1 / (1 + getME(m0, "theta")^(-2)))

## school.(Intercept) 
##          0.5768188

So the ICC is quite large for this data set. However, it is important to also quantify the uncertainty of a point estimate. Although there are analytic methods to obtain SE and CI for ICC, a reliable alternative is to do bootstrapping. We first define the function for computing the test statistic:

icc <- function(x) 1 / (1 + x@theta^(-2))
icc(m0)

## [1] 0.5768188

Note that I used the @ operator for faster extraction. With the bootmlm package we can perform various bootstrap methods using the bootstrap_mer() function.

Parametric Bootstrap

We can run parametric bootstrap, which essentially call the lme4::bootMer() function. It’s usually recommended to have a large number of bootstrap samples ( $R$ ), especially for CIs with higher confidence levels. For illustrative purpose I will use $R = 999$ , but in general 1,999 or more is recommended

system.time(
  boo_par <- bootstrap_mer(m0, icc, nsim = 999L, type = "parametric")
)
boo_par

## 
## PARAMETRIC BOOTSTRAP
## 
## 
## Call:
## lme4::bootMer(x = x, FUN = FUN, nsim = nsim, seed = seed, use.u = FALSE, 
##     type = "parametric", verbose = FALSE)
## 
## 
## Bootstrap Statistics :
##      original      bias    std. error
## t1* 0.5768188 -0.01019232  0.08598194

As you can see, the SE for the ICC is estimated to be 0.0859819 with parametric bootstrap.

Confidence interval

With parametric bootstrap there are three ways to construct confidence intervals via the boot.ci() function from the boot package: normal, basic, and percentile. We can use the following function:

boo_par_ci <- boot::boot.ci(boo_par, type = c("norm", "basic", "perc"), 
                            index = 1L)
boo_par_ci

## BOOTSTRAP CONFIDENCE INTERVAL CALCULATIONS
## Based on 999 bootstrap replicates
## 
## CALL : 
## boot::boot.ci(boot.out = boo_par, type = c("norm", "basic", "perc"), 
##     index = 1)
## 
## Intervals : 
## Level      Normal              Basic              Percentile     
## 95%   ( 0.4185,  0.7555 )   ( 0.4369,  0.7813 )   ( 0.3723,  0.7167 )  
## Calculations and Intervals on Original Scale

Residual Bootstraps

Whereas parametric bootstrap resamples from independent normal distributions, residual bootstrap samples the residuals. Therefore, residual bootstrap is expected to be more robust to non-normality. bootmlm implements three methods for residual bootstrap: differentially reflated residual bootstrap, Carpenter-Goldstein-Rashbash’s residual bootstrap (CGR; Carpenter et al., 2003), and transformational residual bootstrap by van der Leeden, Meijer, and Busin (2008). They are all motivated by the fact that the residuals, generally empirical bayes estimates (denoted as $\tilde u$ and $\tilde e$ ), are shrinkage estimates and have sampling variabilities much smaller than the population random effects, $u$ and $e$ .

The first residual bootstrap rescale $\tilde u$ and $\tilde e$ so that their sampling variabilities match those of $u$ and $e$ as implied by the model estimates. This can be obtained by

system.time(
  boo_res <- bootstrap_mer(m0, icc, nsim = 999L, type = "residual")
)
boo_res

## 
## ORDINARY NONPARAMETRIC BOOTSTRAP
## 
## 
## Call:
## bootstrap_mer(x = m0, FUN = icc, nsim = 999, type = "residual")
## 
## 
## Bootstrap Statistics :
##      original       bias    std. error
## t1* 0.5768188 -0.008674732  0.07221562

As you can see, the SE for the ICC is estimated to be 0.0722156 with residual bootstrap.

The second method, CGR, rescale the sample covariance matrix of the realized values of the residuals to match the model-implied variance components. This can be obtained by

system.time(
  boo_cgr <- bootstrap_mer(m0, icc, nsim = 999L, type = "residual_cgr")
)
boo_cgr

## 
## ORDINARY NONPARAMETRIC BOOTSTRAP
## 
## 
## Call:
## bootstrap_mer(x = m0, FUN = icc, nsim = 999, type = "residual_cgr")
## 
## 
## Bootstrap Statistics :
##      original       bias    std. error
## t1* 0.5768188 -0.007783322  0.07040576

The SE is estimated to be 0.0704058 with CGR bootstrap.

The third method first transforms the OLS residuals, $\hat r_{ij} = y_{ij} - \boldsymbol{x}_{ij} \hat{\boldsymbol{\beta}}$ , by the inverse of cholesky factor, $\boldsymbol{L}$ , of the model-implied covariance matrix of $\boldsymbol{y}$ , $\hat{\boldsymbol{V}}$ , so that theoretically $\boldsymbol{L}^{-1} (\boldsymbol{y} - \boldsymbol{X \beta})$ should be independent and identically distributed. However, as the true sampling variance of $\hat r_{ij}$ is not $\boldsymbol{V}$ , I also provide the option corrected_trans = TRUE to do the transformation using the theoretically sampling variability of $\hat r_{ij}$ .

# Transformation according to V
system.time(
  boo_tra <- bootstrap_mer(m0, icc, nsim = 999L, type = "residual_trans")
)
boo_tra

# Transformation according to the sampling variance of r
system.time(
  boo_trac <- bootstrap_mer(m0, icc, nsim = 999L, type = "residual_trans", 
                            corrected_trans = TRUE)
)
boo_trac

## 
## ORDINARY NONPARAMETRIC BOOTSTRAP
## 
## 
## Call:
## bootstrap_mer(x = m0, FUN = icc, nsim = 999, type = "residual_trans")
## 
## 
## Bootstrap Statistics :
##      original       bias    std. error
## t1* 0.5768188 -0.009255973  0.07836078

## 
## ORDINARY NONPARAMETRIC BOOTSTRAP
## 
## 
## Call:
## bootstrap_mer(x = m0, FUN = icc, nsim = 999, type = "residual_trans", 
##     corrected_trans = TRUE)
## 
## 
## Bootstrap Statistics :
##      original      bias    std. error
## t1* 0.5768188 -0.01245717  0.08085008

The SE is estimated to be 0.0783608 and 0.0808501 with and without corrections with the transformational residual bootstrap.

Confidence interval

With residual bootstrap methods there are four ways to construct confidence intervals via the boot.ci() function from the boot package, with the addition of the bias-corrected and accelarted bootstrap (BCa). We can use the following function:

# First need to compute the influence values
inf_val <- empinf_mer(m0, icc, index = 1)

# Residual bootstrap
boo_res_ci <- boot::boot.ci(boo_res, type = c("norm", "basic", "perc", "bca"), 
                            index = 1L, L = inf_val)
boo_res_ci

## BOOTSTRAP CONFIDENCE INTERVAL CALCULATIONS
## Based on 999 bootstrap replicates
## 
## CALL : 
## boot::boot.ci(boot.out = boo_res, type = c("norm", "basic", "perc", 
##     "bca"), index = 1, L = inf_val)
## 
## Intervals : 
## Level      Normal              Basic         
## 95%   ( 0.4440,  0.7270 )   ( 0.4583,  0.7456 )  
## 
## Level     Percentile            BCa          
## 95%   ( 0.4080,  0.6953 )   ( 0.4465,  0.7146 )  
## Calculations and Intervals on Original Scale
## Some BCa intervals may be unstable

# CGR
boo_cgr_ci <- boot::boot.ci(boo_cgr, type = c("norm", "basic", "perc", "bca"), 
                            index = 1L, L = inf_val)
boo_cgr_ci

## BOOTSTRAP CONFIDENCE INTERVAL CALCULATIONS
## Based on 999 bootstrap replicates
## 
## CALL : 
## boot::boot.ci(boot.out = boo_cgr, type = c("norm", "basic", "perc", 
##     "bca"), index = 1, L = inf_val)
## 
## Intervals : 
## Level      Normal              Basic         
## 95%   ( 0.4466,  0.7226 )   ( 0.4538,  0.7281 )  
## 
## Level     Percentile            BCa          
## 95%   ( 0.4255,  0.6998 )   ( 0.4472,  0.7244 )  
## Calculations and Intervals on Original Scale
## Some BCa intervals may be unstable

# Transformational (no correction)
boo_tra_ci <- boot::boot.ci(boo_tra, type = c("norm", "basic", "perc", "bca"), 
                            index = 1L, L = inf_val)
boo_tra_ci

## BOOTSTRAP CONFIDENCE INTERVAL CALCULATIONS
## Based on 999 bootstrap replicates
## 
## CALL : 
## boot::boot.ci(boot.out = boo_tra, type = c("norm", "basic", "perc", 
##     "bca"), index = 1, L = inf_val)
## 
## Intervals : 
## Level      Normal              Basic         
## 95%   ( 0.4325,  0.7397 )   ( 0.4570,  0.7664 )  
## 
## Level     Percentile            BCa          
## 95%   ( 0.3872,  0.6966 )   ( 0.4241,  0.7081 )  
## Calculations and Intervals on Original Scale

# Transformational (with correction)
boo_trac_ci <- boot::boot.ci(boo_trac, type = c("norm", "basic", "perc", "bca"), 
                             index = 1L, L = inf_val)
boo_trac_ci

## BOOTSTRAP CONFIDENCE INTERVAL CALCULATIONS
## Based on 999 bootstrap replicates
## 
## CALL : 
## boot::boot.ci(boot.out = boo_trac, type = c("norm", "basic", 
##     "perc", "bca"), index = 1, L = inf_val)
## 
## Intervals : 
## Level      Normal              Basic         
## 95%   ( 0.4308,  0.7477 )   ( 0.4498,  0.7624 )  
## 
## Level     Percentile            BCa          
## 95%   ( 0.3913,  0.7038 )   ( 0.4229,  0.7331 )  
## Calculations and Intervals on Original Scale
## Some BCa intervals may be unstable

Random Effect Block Bootstrap

system.time(
  boo_reb <- bootstrap_mer(m0, icc, nsim = 999L, type = "reb")
)
boo_reb

system.time(
  boo_rebs <- bootstrap_mer(m0, icc, nsim = 999L, type = "reb", 
                            reb_scale = TRUE)
)
boo_rebs

## 
## ORDINARY NONPARAMETRIC BOOTSTRAP
## 
## 
## Call:
## bootstrap_mer(x = m0, FUN = icc, nsim = 999, type = "reb")
## 
## 
## Bootstrap Statistics :
##      original     bias    std. error
## t1* 0.5768188 0.06957508  0.06176538

## 
## ORDINARY NONPARAMETRIC BOOTSTRAP
## 
## 
## Call:
## bootstrap_mer(x = m0, FUN = icc, nsim = 999, type = "reb", reb_scale = TRUE)
## 
## 
## Bootstrap Statistics :
##      original      bias    std. error
## t1* 0.5768188 -0.01103587  0.07066787

Confidence interval

# Only sampling clusters
boo_reb_ci <- boot::boot.ci(boo_reb, type = c("norm", "basic", "perc", "bca"), 
                            index = 1L, L = inf_val)

## Warning in norm.inter(t, adj.alpha): extreme order statistics used as endpoints

boo_reb_ci

## BOOTSTRAP CONFIDENCE INTERVAL CALCULATIONS
## Based on 999 bootstrap replicates
## 
## CALL : 
## boot::boot.ci(boot.out = boo_reb, type = c("norm", "basic", "perc", 
##     "bca"), index = 1, L = inf_val)
## 
## Intervals : 
## Level      Normal              Basic         
## 95%   ( 0.3862,  0.6283 )   ( 0.4026,  0.6408 )  
## 
## Level     Percentile            BCa          
## 95%   ( 0.5129,  0.7510 )   ( 0.3551,  0.6389 )  
## Calculations and Intervals on Original Scale
## Warning : BCa Intervals used Extreme Quantiles
## Some BCa intervals may be unstable

# Transformational (with correction)
boo_rebs_ci <- boot::boot.ci(boo_rebs, type = c("norm", "basic", "perc", "bca"), 
                             index = 1L, L = inf_val)
boo_rebs_ci

## BOOTSTRAP CONFIDENCE INTERVAL CALCULATIONS
## Based on 999 bootstrap replicates
## 
## CALL : 
## boot::boot.ci(boot.out = boo_rebs, type = c("norm", "basic", 
##     "perc", "bca"), index = 1, L = inf_val)
## 
## Intervals : 
## Level      Normal              Basic         
## 95%   ( 0.4493,  0.7264 )   ( 0.4642,  0.7419 )  
## 
## Level     Percentile            BCa          
## 95%   ( 0.4117,  0.6895 )   ( 0.4521,  0.7157 )  
## Calculations and Intervals on Original Scale
## Some BCa intervals may be unstable

Case Bootstrap

With case bootstrap, the observed cases are sampled with replacement. However, because of the multilevel structure, we need to resample the clusters. Optionally, we can then resample the cases within each cluster (using the lv1_resample = TRUE argument). Unlike the parametric and residual bootstrap methods, currently bootmlm only support the case bootstrap with two levels.

system.time(
  boo_cas <- bootstrap_mer(m0, icc, nsim = 999L, type = "case")
)
boo_cas

system.time(
  boo_cas1 <- bootstrap_mer(m0, icc, nsim = 999L, type = "case", 
                            lv1_resample = TRUE)
)
boo_cas1

## 
## ORDINARY NONPARAMETRIC BOOTSTRAP
## 
## 
## Call:
## bootstrap_mer(x = m0, FUN = icc, nsim = 999, type = "case")
## 
## 
## Bootstrap Statistics :
##      original      bias    std. error
## t1* 0.5768188 -0.01577482  0.05671946

## 
## ORDINARY NONPARAMETRIC BOOTSTRAP
## 
## 
## Call:
## bootstrap_mer(x = m0, FUN = icc, nsim = 999, type = "case", lv1_resample = TRUE)
## 
## 
## Bootstrap Statistics :
##      original     bias    std. error
## t1* 0.5768188 0.06758788  0.06112113

The SE for the ICC is estimated to be 0.0567195 (only sampling clusters) and 0.0611211 (sampling also cases) with case bootstrap.

Confidence interval

With case bootstrap the supported CIs are: normal, basic, and percentile, and BCa. We can use the following function:

# Only sampling clusters
boo_cas_ci <- boot::boot.ci(boo_cas, type = c("norm", "basic", "perc", "bca"), 
                            index = 1L, L = inf_val)
boo_cas_ci

## BOOTSTRAP CONFIDENCE INTERVAL CALCULATIONS
## Based on 999 bootstrap replicates
## 
## CALL : 
## boot::boot.ci(boot.out = boo_cas, type = c("norm", "basic", "perc", 
##     "bca"), index = 1, L = inf_val)
## 
## Intervals : 
## Level      Normal              Basic         
## 95%   ( 0.4814,  0.7038 )   ( 0.4953,  0.7223 )  
## 
## Level     Percentile            BCa          
## 95%   ( 0.4314,  0.6583 )   ( 0.4807,  0.6838 )  
## Calculations and Intervals on Original Scale
## Some BCa intervals may be unstable

# Transformational (with correction)
boo_cas1_ci <- boot::boot.ci(boo_cas1, type = c("norm", "basic", "perc", "bca"), 
                             index = 1L, L = inf_val)

## Warning in norm.inter(t, adj.alpha): extreme order statistics used as endpoints

boo_cas1_ci

## BOOTSTRAP CONFIDENCE INTERVAL CALCULATIONS
## Based on 999 bootstrap replicates
## 
## CALL : 
## boot::boot.ci(boot.out = boo_cas1, type = c("norm", "basic", 
##     "perc", "bca"), index = 1, L = inf_val)
## 
## Intervals : 
## Level      Normal              Basic         
## 95%   ( 0.3894,  0.6290 )   ( 0.4008,  0.6435 )  
## 
## Level     Percentile            BCa          
## 95%   ( 0.5101,  0.7529 )   ( 0.4189,  0.6407 )  
## Calculations and Intervals on Original Scale
## Warning : BCa Intervals used Extreme Quantiles
## Some BCa intervals may be unstable

Summary

boo_names <- c("parametric", "residual", "cgr", "trans", 
               "trans (cor)", "REB", "REB (scaled)", 
               "case (cluster)", "case (c + i)")
boo_lst <- list(boo_par, boo_res, boo_cgr, boo_tra, boo_trac, 
                boo_reb, boo_rebs, boo_cas, boo_cas1)
boo_ci_lst <- list(boo_par_ci, boo_res_ci, boo_cgr_ci, boo_tra_ci, 
                   boo_trac_ci, boo_reb_ci, boo_rebs_ci, boo_cas_ci, 
                   boo_cas1_ci)

get_ci <- function(boo_ci, type) {
  paste0("(", paste(comma(tail(boo_ci[[type]][1, ], 2L)), collapse = ", "), ")")
}

tab <- tibble(boot_type = boo_names, boo = boo_lst, boo_ci = boo_ci_lst) |>
  mutate(sd = sapply(boo, \(x) comma(sd(x$t))), 
         normal = sapply(boo_ci, \(x) get_ci(x, "normal")), 
         basic = sapply(boo_ci, \(x) get_ci(x, "basic")), 
         percentile = sapply(boo_ci, \(x) get_ci(x, "percent")), 
         bca = sapply(boo_ci, \(x) get_ci(x, "bca"))) |>
  select(-boo, -boo_ci)

knitr::kable(tab)

boot_type	sd	normal	basic	percentile	bca
parametric	0.086	(0.42, 0.76)	(0.44, 0.78)	(0.37, 0.72)	(NULL)
residual	0.072	(0.44, 0.73)	(0.46, 0.75)	(0.41, 0.70)	(0.45, 0.71)
cgr	0.07	(0.45, 0.72)	(0.45, 0.73)	(0.43, 0.70)	(0.45, 0.72)
trans	0.078	(0.43, 0.74)	(0.46, 0.77)	(0.39, 0.70)	(0.42, 0.71)
trans (cor)	0.081	(0.43, 0.75)	(0.45, 0.76)	(0.39, 0.70)	(0.42, 0.73)
REB	0.062	(0.39, 0.63)	(0.40, 0.64)	(0.51, 0.75)	(0.36, 0.64)
REB (scaled)	0.071	(0.45, 0.73)	(0.46, 0.74)	(0.41, 0.69)	(0.45, 0.72)
case (cluster)	0.057	(0.48, 0.70)	(0.50, 0.72)	(0.43, 0.66)	(0.48, 0.68)
case (c + i)	0.061	(0.39, 0.63)	(0.40, 0.64)	(0.51, 0.75)	(0.42, 0.64)

Mark Lai

2026-05-05

Data

Parametric Bootstrap

Confidence interval

Residual Bootstraps

Confidence interval

Random Effect Block Bootstrap

Confidence interval

Case Bootstrap

Confidence interval

Summary